Postprandial glucose excursions can inhibit achievement of good glycaemic control, and possibly have a specific effect on the risk of vascular comorbidities. Rapid-acting analogues control these excursions better than human insulin because their pharmacokinetic/pharmacodynamic (PK/PD) profile is closer to that of meal-time endogenous insulin secretion. Review of the findings of PK/PD studies and clinical trials suggests that the three marketed rapid-acting analogues—insulin lispro, insulin aspart and insulin glulisine—are equally efficacious and safe. In comparison with human insulin when using the same basal insulin, they provide comparable glycaemic control with a reduced risk of hypoglycaemia, although the combination of rapid-acting and basal analogues reduces glycated haemoglobin (HbA1c) more than human meal-time insulin combined with neutral protamine Hagedorn (NPH) insulin. Some studies have suggested that insulin glulisine has a slightly faster onset of action compared with insulin lispro or insulin aspart, but this has not been translated into demonstrable clinical benefit. Treatment satisfaction in patients with diabetes has been higher when therapy with a rapid-acting analogue is used instead of human insulin, perhaps due to differences in advised timing of injection. The largest benefits in efficacy, hypoglycaemia incidence, treatment satisfaction and quality of life have occurred when patients receive an all-analogue meal-time plus basal regimen as compared with an all-human insulin regimen. No new safety issues have been identified with the marketed rapid-acting analogues, and their insulin-like growth factor 1receptor affinity and mitogenic activity are comparable to human insulin.